Comment by tharmas

Does this mean us plebs can run LLMs on gimped VRAM Nvidia lower end cards?